Init First NanoClaw Agent
Guides you through creating your first NanoClaw DM agent and verifies delivery with a welcome message.
AutomationAdvanced★ 30,471⑂ 12,865AI score 6/10Last updated: Aug 9, 2026
What it does
This skill walks an operator through standing up the first agent in NanoClaw, a multi-channel messaging agent runtime.
- Reads
src/channels/index.tsand.envto list enabled channels and asks which one should host the DM - Collects your channel user ID, display name, and the agent persona name
- Resolves the DM
platform_id: direct-addressable channels (Telegram, WhatsApp, iMessage, Matrix, Resend) reuse the user handle, while cold-DM channels (Discord, Slack, Teams, Webex, gChat) require you to DM the bot once so the router records amessaging_groupsrow - Runs
scripts/init-first-agent.tsto upsert the user, create the agent group and its filesystem, and wire messaging group ↔ agent bindings - Sends a welcome DM through the service CLI socket for a true end-to-end test, with an
inbound.dbfallback - Ships a concrete troubleshooting section (stuck
pendingrows, ACL rejections, unknown-sender policy, name collisions)
Every step is idempotent, so reruns are safe.
Who it's for
- Operators who just installed NanoClaw and need their first working agent
- Anyone confused by the plumbing between a chat bot and an agent group
- Users who need a systematic checklist when the welcome DM never arrives
Note: only meaningful inside the NanoClaw repo, with the service running and channel credentials already configured.
Examples
- First Discord agent: after
/add-discord, invoke the skill, DM the bot once, confirm the returnedplatform_id, and receive the welcome DM. - Telegram pair-code path: if you'd rather not DM first, use
--step pair-telegram, send the issued code in Telegram, and the interceptor links your user automatically. - Delivery debugging: nothing arrives in two minutes → query
outbound.dbfor pending rows and grep the log forUnauthorized channel destinationor container exits.
· · · Install guide · · ·
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on Claude Skill Mart. Copy the .claude/skills/init-first-agent folder from the GitHub repo nanocoai/nanoclaw into my ~/.claude/skills/init-first-agent/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/nanocoai/nanoclaw.git && mkdir -p ~/.claude/skills && cp -r nanoclaw/.claude/skills/init-first-agent ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal and clone the repo:
git clone https://github.com/nanocoai/nanoclaw.git - Create the skills directory if needed:
mkdir -p ~/.claude/skills - Copy the skill:
cp -r nanoclaw/.claude/skills/init-first-agent ~/.claude/skills/- (If you work directly inside the NanoClaw repo, the skill is already available — no copy needed.)
- Set up NanoClaw first by running
/setup, then confirm the service is running under launchd (macOS) or systemd (Linux). - Run a channel skill (
/add-discord,/add-telegram, …) so credentials land in.envand the adapter is uncommented insrc/channels/index.ts. - Launch Claude Code from the NanoClaw repo directory and ask it to "create my first agent" or invoke
init-first-agent. - Wait for the welcome DM (first container cold start takes ~60s); if it doesn't arrive, follow the diagnostic commands the skill prints.
View source on GitHub ↗License: MIT